Image and video compression is becoming an increasing important area of investigation, with numerous applications to video conferencing, interactive education, home entertainment, and potential application to earth observation, medical imaging, digital libraries and many other areas. In this paper the Kohene's self-organizing feature map (KSOFM) neural network and a modified frequency, sensitive-competitive learning algorithm have been utilized with a great deal of success to overcome the problem of codebook design in vector quantization. A detailed investigation of the network parameters has been conducted to achieve the optimal performance in image compression. A set of fourteen color images obtained from the internet were used as training and test samples. The results have shown that the quality of the decompressed images, was compared to the originals visually and using the peak-signal-to-noise-ratio (PSNR) as a measure of quality.
Soliman, H., Awadalla, M., & Abou Chadi, F. (2021). Self-Organizing Map for Image Compression: A Study of Optimal Performance.. MEJ- Mansoura Engineering Journal, 24(2), 27-36. doi: 10.21608/bfemu.1999.147596
MLA
H. H. Soliman; M. M. Awadalla; F. E. Z. Abou Chadi. "Self-Organizing Map for Image Compression: A Study of Optimal Performance.". MEJ- Mansoura Engineering Journal, 24, 2, 2021, 27-36. doi: 10.21608/bfemu.1999.147596
HARVARD
Soliman, H., Awadalla, M., Abou Chadi, F. (2021). 'Self-Organizing Map for Image Compression: A Study of Optimal Performance.', MEJ- Mansoura Engineering Journal, 24(2), pp. 27-36. doi: 10.21608/bfemu.1999.147596
VANCOUVER
Soliman, H., Awadalla, M., Abou Chadi, F. Self-Organizing Map for Image Compression: A Study of Optimal Performance.. MEJ- Mansoura Engineering Journal, 2021; 24(2): 27-36. doi: 10.21608/bfemu.1999.147596
)
View on SCiNiTO